Hello, I'm Bea and I’m looking for a postdoc position! So far, I've worked on multi-agent systems (MSc) and on automating crowd simulation tasks (PhD). I love teaching and programming and I solve most of my problems with questionable AIs in Python.
2017 – 2022. PhD in Information and Communication Technologies (Universitat Pompeu Fabra).
2016 – 2017. Master in Intelligent Interactive Systems (Universitat Pompeu Fabra).
2012 – 2016. Bachelor’s degree in Computer Science (Universitat Pompeu Fabra).
2010 – 2012. Bachibac Program. Double award Batxillerat – Baccalauréat (IES Vicenç Plantada).
Teaching (Universitat Pompeu Fabra), 2017-2022.
Summer School course instructor (Campus Junior), 2020 - 2021.
Robotics instructor (American Spaces), January 2018 – June 2018.
Research assistant (Universitat Pompeu Fabra), December 2016 – September 2017.
Java Developer Intern (everis), April 2016 – September 2016.
Summer Intern (Teixidó SA), July 2012 – August 2012.
Home tutor, January 2009 – August 2015.
I volunteer, sometimes to raise funds, for different causes e.g. Red Cross, La Marató de TV3 (solidarity project that focuses on a different disease each year), Oracle4Girls, Girls In Tech, Mobile Social Congress, etc.
Python
C/C++/C#
JavaScript
CSS
Matlab
Java
GLSL
LaTeX
Robotic Operative System (ROS)
Scipy
Keras
Three.js
Adobe Premiere
Google Docs
Matlab
Adobe Photoshop
Sony Vegas
Google Docs
OBS
DaVinci Resolve
Catalan. Native or bilingual proficiency.
Spanish. Native or bilingual proficiency.
English. British Council, CEFR level C2.
French. Alliance Française, DELF niveau B2.
German. Escola Oficial d’Idiomes, CEFR A2.
2022. Dynamic Combination of Crowd Steering Policies Based on Context. To be published in special issue of Computer Graphics Forum, Eurographics'22 conference proceedings. In this paper, we study the performance of a number of steering policies (i.e., crowd simulation algorithm and its parameters) in a variety of contexts, resorting to an existing quality function able to automatically evaluate simulation results. This analysis allows us to map contexts to the performance of steering policies. Furthermore, we propose a solution to dynamically adjust the policies based on the local context each agent is currently in.
2021. A Perceptually-Validated Metric for Crowd Trajectory Quality Evaluation. https://dl.acm.org/doi/abs/10.1145/3480136. In this paper we study the relation between parametric values for simulation techniques and the quality of the resulting trajectories. A quality metric, QF, is proposed to abstract from reference data while capturing the most salient features that affect the perception of trajectory realism. These trajectory features are selected from the literature and from interviews with experts. To validate the capacity of QF to capture perceived trajectory quality, we conduct an online experiment that demonstrates the high agreement between the automatic quality score and non-expert users.
2020. Generalised Microscopic Crowd Simulation using Costs in Velocity Space. https://doi.org/10.1145/3384382.3384532. Many algorithms to simulate human crowd behaviour have been proposed, each using different principles and implementation details that are difficult to compare. This paper presents a novel framework that describes local agent navigation generically as optimising a cost function in a velocity space. We show that many state-of-the-art algorithms can be translated to this framework using a single general principle. This software enables easy experimentation with different algorithms and parameters and honest comparisons between them. Our implementation is freely available online.
2017. Cross-Entropy method for Kullback-Leibler control in multi-agent systems. http://hdl.handle.net/10230/33109. We consider the problem of computing optimal control policies in large-scale multiagent systems, for which the standard approach via the Bellman equation is intractable. Our formulation is based on the Kullback-Leibler control framework, also known as Linearly- Solvable Markov Decision Problems. In this setting, adaptive importance sampling methods have been derived that, when combined with function approximation, can be effective for high-dimensional systems.
Course Certificates Completed: Ideas from the History of Graphic Design, Fundamentals of Graphic Design, Introduction to Typography, Introduction to Imagemaking, and Brand New Brand.
Course info CertificationI fought very hard for this one but I enjoyed every second of it.
Course info Certification3D graphics and printing.
Communication.
Content creation.